Font Size: a A A

Study On Choke Pressure Control Technology In Managed Pressure Drilling Based On Neural Networks

Posted on:2015-05-27Degree:MasterType:Thesis
Country:ChinaCandidate:Y WangFull Text:PDF
GTID:2271330503455878Subject:Oil-Gas Well Engineering
Abstract/Summary:PDF Full Text Request
Managed pressure drilling provides an effective means of safe drilling in narrow density pressure window. Automatic control system is an important part of managed pressure drilling system. The choke pressure control technology is the key to the automatic control system, and also the difficult part. For this reason, a study on the choke pressure control technology is significant to managed pressure drilling.Based on describing the principle of choke pressure control in this article, which includes casing pressure control and standpipe pressure control, the composition of choke pressure system is introduced. Studies have shown that: since the control loop in casing pressure control is short, the adjustment time is very short, it is relatively easy to control; since the control loop in standpipe pressure control is very long, there is large time delay, so the adjustment time is very long, it is relatively hard to control. Therefore, this paper focuses on the design of the choke pressure control program.By comparing mainstream control technology, combining with controlled pressure drilling characteristics, artificial neural network technology is preferred as a method of controlling the choke pressure control. Dual-network control system for choke pressure control is constructed. The three components of control system-controller, identifier, predictor – is designed. The stability of system is analyzed. The system is able to effectively solve the problems such as nonlinear, large time delay, inaccuracy of mathematical model.In this paper, the simulation model, including choke and electro-hydraulic servo system,is established, the implementation steps of the simulation program is established, simulation is implemented for different input target. Simulation results show the good results of dual network control system.
Keywords/Search Tags:Managed pressure drilling, Choke pressure, Neural networks, Nonlinear, Time delay
PDF Full Text Request
Related items